tribological behaviour of pulsed magnetron sputtered crb2 coatings examined by reciprocating sliding...

Post on 26-Nov-2023

0 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

Available online at www.sciencedirect.com

202 (2008) 1470–1478www.elsevier.com/locate/surfcoat

Surface & Coatings Technology

Tribological behaviour of pulsed magnetron sputtered CrB2 coatingsexamined by reciprocating sliding wear testing against

aluminium alloy and steel

M. Audronis a,⁎, Z.M. Rosli a, A. Leyland a, P.J. Kelly b, A. Matthews a

a Department of Engineering Materials, The University of Sheffield, Sir Robert Hadfield Building, Mappin Street, Sheffield, S1 3JD, UKb Surface Engineering Group, Manchester Metropolitan University, Manchester, M1 5GD, UK

Received 7 April 2007; accepted in revised form 28 June 2007Available online 5 July 2007

Abstract

Among the number of attractive properties that transition-metal diborides (TiB2, CrB2, etc.) possess, high resistance to wear and chemicalinertness are themost important when considering diboride coatings for dry machining of nonferrous materials, such as aluminium and its alloys. Duemostly to the problematic deposition of chromium diboride (preparation of targets, target cracking during the deposition process, control ofstoichiometry etc.), these coatings remain comparatively less studied than, for example, titanium diborides, regarding their tribological performance.

In this paper we report on the tribological behaviour of pulsed magnetron sputtered (PMS), smooth and fully dense, crystalline, 21–38 GPahard CrB2 coatings examined by reciprocating sliding wear testing in ambient air (20±2 °C, 20–30% humidity) against EN AW-2017Aaluminium alloy and AISI 52100 chrome steel. The results are compared to those of pulsed magnetron sputter deposited TiN and CrN coatings. Itis demonstrated that pulsed magnetron sputtered chromium diboride coatings exhibit the best tribological performance, in terms of amount ofaluminium adhered on the surface of the wear track, during testing against aluminium alloy. When slid against AISI 52100 steel PMS CrB2, CrNand TiN coatings exhibited coefficients of friction of 0.6, 0.6–0.7 and 0.43–0.45 respectively. The tribological behaviour of coatings was found tobe dependent on the transfer film formation and its properties. Wear rates were up to ten times lower for pulsed magnetron sputtered CrB2

coatings, compared to DC sputtered Cr–B films.© 2007 Elsevier B.V. All rights reserved.

Keywords: Chromium diboride; Sliding wear; Aluminium; Steel; Pulsed magnetron sputtering; Titanium and chromium nitride

1. Introduction

Among the number of attractive properties that transition-metal diborides possess, high resistance to wear and chemicalinertness are the most important when considering diboridecoatings for dry machining of nonferrous materials, such asaluminium and its alloys [1–6]. However, the deposition ofchromium diboride is quite problematic (in terms of preparationof targets, target cracking during the deposition process, control ofstoichiometry etc. [7,8]) and therefore these coatings remain

⁎ Corresponding author. Tel.: +44 114 2225968; fax: +44 114 2225943.E-mail address: m.audronis@yahoo.co.uk (M. Audronis).

0257-8972/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.surfcoat.2007.06.057

comparatively less studied than, for example, titanium diborides[1-6,9,10], regarding their tribological performance.

Recently, CrB2 coatings were successfully depositedemploying a new deposition method which involves sputteringof loosely packed powder targets [11,12]. Sputtering of looselypacked blended boron and chromium powder targets in mediumfrequency asymmetric bipolar pulsed-DC mode [13] ensuredthe deposition of hard (H in a range of 21 to 38 GPa) andcrystalline CrB2 coatings possessing smooth surfaces[11,12,14,15]. The harder coatings (H N30 GPa) were shownto possess a nano-columnar structure and higher degree of (001)texture as compared to their less hard counterparts [16].

The aim of this paper was to investigate the tribologicalbehaviour of pulsed magnetron sputtered CrB2 coatingsemploying the reciprocating sliding wear testing against EN

Table 1Properties of Cr–B, TiN and CrN coatings

Coating Hardness H, GPa Average roughness Ra, nm Thickness, μm Structure Phase composition

PMS CrB2 21–38 10–25 1.1–1.9 Crystalline, Zone II/III CrB2 [11,16]DC Cr–B (Cr:B — 1:3) – a ~900 ~1.4 Amorphous, Zone T Amorphous phase+CrB2 crystallites [16,21]PSM TiN 30 20 1.9 Crystalline, Zone II/III TiNPMS CrN 18 10 1.4 Crystalline, Zone II/III CrN, Cr2Na It was not possible to measure hardness of this sample correctly due to very high roughness of DC Cr–B coatings.

1471M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

AW-2017A aluminium alloy and AISI 52100 steel. It was alsoan objective to compare the behaviour of PMS CrB2 coatings tothat of DC sputtered Cr–B and PMS TiN and CrN coatings.

2. Experimental procedure

CrB2 coatings were deposited in a magnetron sputterdeposition rig with closed field unbalanced magnetron(CFUBM) configuration [12]. The rig contained a 180 mmdiameter unbalanced magnetron, installed in the ‘sputter-up’configuration. The substrate holder, mounted 100 mm above thetarget, was water-cooled to ensure constant substrate temperatureprofiles during each run. Substrate temperature (monitored by aK-type thermocouple attached to the substrate holder) was in therange of 110–150 °C during all runs. This range of temperaturescorresponds to a homologous temperature Ts/Tm of ~0.2 (meltingpoint of CrB2 is 2123 K), where Ts is the substrate temperatureand Tm is the melting point of the coating material.

Fig. 1. SEM micrograph of the fracture section of (a) a CrB2 coating deposited by pusputtering; (c) a TiN coating deposited by pulsed-DC magnetron sputtering; (d) a C

The powder sputter target consisted of Cr (99.99% purity)and amorphous B powders blended at an atomic ratio of 1:2. Forboth powders the average particle size was ~5 micron, and theblend was mixed in a rotating drum for several hours. Nocompaction was performed after mixing the blend. Additionalpieces of solid boron (approximately 1 cm3 each) were alsoplaced at regular intervals around the ‘racetrack’ region toachieve the required chromium diboride film stoichiometry [12].The target power, delivered by a dual channel (5+5 kW)Advanced Energy Industries ‘Pinnacle Plus’ unit, was set to500W during each run. The target pulsing frequency was kept ata constant 100 kHz value while the pulse duty cycle was variedin a range of 60–80%. Substrates were left floating, or biasednegatively to either −30 or −85 V. For the purposes ofcomparison, Cr–B coatings were also deposited by DC sputterdeposition, under otherwise identical process conditions.

TiN and CrN coatings were also deposited by sputter depositionusing an unbalanced magnetron configuration. In this case the

lsed-DC magnetron sputtering; (b) a Cr–B coating deposited by DC magnetronrN coating deposited by pulsed-DC magnetron sputtering.

Fig. 2. Coefficient of friction for PMS CrB2 ((1) did not fail, (2) failed duringtesting), TiN, and CrN coatings tested against Ø10 mm EN AW-2017Aaluminium alloy ball at 5 N normal load.

1472 M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

chamber contained a vertically installed rectangular unbalancedmagnetron fitted with a 380 mm×100 mm size 99.99% pure Ti orCr target. The substrate holder was mounted 140 mm away fromthe target. The coatings were deposited onto substrates mounted ona stainless steel substrate holder with no external heating applied.The power applied to the target was 1200 W in both cases. ENIRPG-100 and ENIDCG-100 10 kWpower supplies were used as atarget sputter and substrate bias power sources, respectively. Thedeposition runs were carried out at a target pulsing frequency of130 kHz and 72% duty cycle and the substrates were biasednegatively to −50 V. The coatings were deposited at substratetemperatures of 100±10 °C. This substrate temperature wouldcorrespond to Ts/Tm of 0.11 and 0.21 for TiN (melting point —3200 K) and CrN (melting point — 1770 K) respectively. Thedeposition time was 74 min for TiN and 37 min for CrN. The flowof nitrogen was controlled by a reactive sputtering processcontroller (‘Reactaflo') equipped with an optical emissionmonitoring (OEM) system, which was tuned to the 365 and

Fig. 3. A comparison of friction curves for DC and pulsed-DC deposited Cr–Bcoatings. Coatings were tested against Ø10 mm ENAW-2017A aluminium alloyball at 5 N normal load for 250 cycles.

359 nm lines of the Ti and Cr emission spectra, respectively. AnOEM signal of 60% for TiN and 73% for CrN was used fordeposition (i.e., reactive gas was allowed into the chamber until theOEM metal signal had fallen to a preset proportion of the initial100%value). A feedback loop thenmaintained the optical emissionsignal at this proportionally lower value for the duration of thedeposition run.

All coatings investigated in the present study were depositedonto polished silicon (100) wafers and AISI M2 tool steelcoupons (polished to a 1 micron finish, giving an Ra value of~10 nm). The substrates were cleaned ultrasonically inisopropanol followed by 20 min sputter cleaning, at −650 V forCrB2 and −1000 V for TiN and CrN coatings, prior to deposition.

Reciprocating slidingwear testingwas employed to evaluate theCrB2, TiN and CrN coatings deposited on the AISI M2 tool steelsubstrates (Ra ~10 nm; hardness — 50 HRC) against 10 mmdiameter EN AW-2017A aluminium alloy (107 HB) and AISI52100 steel balls (64 HRC). Coatings were tested using the fol-lowing parameters: normal load: 5 N; track length: 10 mm; fre-quency: 5 Hz; sliding speed: 0.1 ms−1; number of cycles: 250 and15,000; atmosphere: ambient air (20±2 °C, 20–30% humidity).

SEM analyses of Cr–B, TiN and CrN coated samples wereperformed using a JEOL 6400microscope operated at 15 kVand aCamScan Mk2 microscope operated at 20 kV. Images wererecorded in both secondary and backscattered electron imaging(SEI and BEI, respectively) modes. Image analysis of BEImicrographs (nine recorded for every sample) was employed tocalculate the area of each wear track that was covered withaluminium, produced by sliding against the aluminium ballcounterface. For the details on contrast generation and proceduresof image analysis, references [17–20] provide further information.

Coating mechanical properties were evaluated by nanoin-dentation of films deposited on Si wafers, using a Micromater-ials Ltd. Nanotest 500 (for aCrB2 coatings) and Hysitron Inc.Triboscope® (for TiN and CrN coatings). A Berkovitch pyramiddiamond indenter was used to make twenty indents per sampleat a constant maximum indentation depth, which was alwaysbelow 10% of the coating thickness.

Fig. 4. Wear track profile scan of CrB2 coating tested against Ø10 mm EN AW-2017A aluminium alloy ball at 5 N normal load.

1473M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

3. Results and discussion

3.1. Morphology, structure andmechanical properties of coatings

The properties of coatings investigated in this paper aresummarised in Table 1. All PMS coatings were crystalline, hardand dense and had smooth surfaces. The morphology of suchcoatings is usually ascribed to the Zone II/III regions of thewidely accepted structure zone models [22–24]. Fig. 1a–dshows SEM micrographs of fracture cross-sections throughPMS CrB2, DC sputtered Cr–B, PMS TiN and PMS CrNcoatings respectively. In contrast to the DC and pulsed-DC Cr–B coatings, TiN and CrN coatings had a thin metallic adhesioninterlayer (known to improve coating wear behaviour [25]). Itwas not possible to deposit adhesion interlayers for the Cr–Bcoatings due to equipment limitations.

3.2. Reciprocating sliding wear testing against EN AW-2017Aaluminium alloy

The friction curves obtained for coatings tested against ENAW-2017A aluminium alloy at 5 N normal load, 0.1 ms−1

sliding speed and at an ambient atmosphere condition (20–30%humidity, 20 °C) are given in Fig. 2. Behaviour of two PMSCrB2 coatings, of which one has failed at about 2000 cycles, iscompared to that of TiN and CrN coatings. The coefficient offriction, plotted against the number of cycles, shows the differentcoating behaviours. For both CrB2 coatings the coefficient of

Fig. 5. SEM SEI (a) & (c) and BEI (b) & (d) mode images of CrB2 coating, tested aglower (a) & (b) and higher (c) & (d) magnifications.

friction (COF) is increasing over the first 750 cycles, from anapproximate value of 0.25 to about 0.55, where it stabilizes and,for coating CrB2 (1), becomes almost constant over theremaining number of cycles. Coating CrB2 (2), however, failsafter about 2000 cycles, followed by a subsequent more rapidtransfer of aluminium and decrease of COF to about 0.4 — atypical COF value of aluminium sliding against aluminium. TheTiN coating exhibits behaviour similar to that of CrB2 (1), i.e. thecoating doesn't fail and the COF is more or less constant at anaverage value of ~0.5. The CrN coating exhibits a COF of ~0.4(the same as failed CrB2 (2) sample) from almost very beginningof testing, which suggests a higher rate of Al transfer and,therefore, relatively worse performance.

The friction behaviour of DC-deposited Cr–B coatingssliding against an aluminium alloy counterface was alsoinvestigated. Fig. 3 compares the friction curve obtained forthe DC Cr–B coating to the curve which was common for thepulsed-DC CrB2 coatings investigated in this study. It appearsfrom the plots that, in terms of friction behaviour, the DCdeposited Cr–B coating exhibits a higher COF through thisinitial period of testing, which is likely to be due to differentwear mechanisms occurring in the wear track, determined bycoating properties such as structure and composition. Fig. 3 alsodemonstrates that pulsed-DC coatings, due to their smooth andfully dense morphology, offer an improved tribologicalperformance and therefore are preferable.

The wear tracks were investigated by surface profilometry.Fig. 4 shows an example of the wear track profile, from which

ainst Ø10 mm EN AW-2017A aluminium alloy ball at 5 N normal load, taken at

Fig. 6. EDX spectra of a) delaminated region (substrate) (‘1’ in Fig. 5d); b) CrB2

coating (‘2’ in Fig. 5d); c) transferred aluminium (‘3’ in Fig. 5d).

Fig. 7. A comparison of wear track area coverage by aluminium (in terms ofpercentage).

1474 M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

no significant wear of the actual coating can be observed. It israther clear from the scan that material has been transferredfrom the ball to the coating. Material transfer has been found tobe the dominant wear mechanism for all Cr–B coatingsinvestigated, as well as for TiN and CrN coatings.

Analysis of wear scars by SEM (Fig. 5a–d) in secondary andbackscattered electron imaging modes has been applied to studythe initial wear mechanisms of CrB2 coatings in reciprocatingsliding against aluminium alloy for two hundred and fifty cycles.EDX has also been used to confirm the chemical composition ofregions of interest. Fig. 5a and b shows plan-view SEI and BEImicrographs of the wear scar. The wear scar morphology anddebris distribution is revealed by SEI imaging (Fig. 5a). The samearea observed in BEI mode (Fig. 5b) distinguishes regionspossessing different chemical composition. Aluminium trans-ferred to the coating surface appears, due to its relatively loweratomic mass, as black areas; the CrB2 coating possesses anintermediate brightness and the substrate material is the brightest.SEMmicrographs shown in Fig. 5c and d demonstrate these threecharacteristic regions at a highermagnification; extra information,such as the presence of cracks in the coating, is also revealed. Thechemical compositions of the regions marked as ‘1’, ‘2’ and ‘3’ in

Fig. 8. Coefficient of friction for PMS CrB2, TiN, and CrN coatings testedagainst Ø10 mm AISI 52100 steel ball at 5 N normal load.

Fig. 9. SEM SEI micrographs showing the features (indicated by arrows) of PMS CrB2 coatings wear scars after 250 cycles; (a) overall view of the wear scar; (b)delaminated regions; (c) steel transferred from the reciprocating sliding ball; (d) fatigue cracks; (e) & (f) roll-shaped particles originating from the transfer film asmaterial is being removed from it.

1475M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

Fig. 5d was confirmed by EDX analysis; the relevant spectra arepresented in a Fig. 6.

It is evident from the analysis performed and the dataobtained that the wear during reciprocating sliding wear testingof CrB2 coatings against aluminium is due to transfer of thelatter from the tip of the ball counterface to the coating surface(‘1’ in Fig. 5d). This is confirmed by the surface profilometryresults shown in Fig. 4. Subsequent reciprocating sliding of aball over the aluminium covered area causes transfer of morematerial on top of the previously adhered aluminium, back-transfer of adhered material to the pin, milling and oxidation oftransferred material and debris, to form a transfer film on thesurface of the coating, and detachment of the previouslytransferred aluminium or material from the oxide-containingtransfer film, forming loose debris. In the case of fatigue crackdevelopment (under cyclic shear forces) at the coating–

substrate interface, localized coating delamination can occur(‘3’ in Fig. 5d). Good adhesion is therefore a determining factorwith regard to coating durability. Delaminated coating materialmilled inside the wear track will affect the composition of weardebris, which will consist of the material that was initiallytransferred to the coating surface (aluminium and aluminiumoxides) and particles from the coating (chromium diboride). Theabrasive nature of aluminium oxide and CrB2 particles willnaturally facilitate abrasive wear of the coating. These particlesare responsible for the grooves seen in Fig. 5a. Similar wearbehaviour has been observed for TiB2 coatings againstaluminium and was reported in references [2,6].

Considering the results discussed above, it is now possible toexplain the friction behaviour of the CrB2 (2) coating. At thebeginning of the sliding wear test (0–500 cycles in Fig. 5) theincrease in friction coefficient is most likely due to increasing

Fig. 10. SEM SEI micrographs showing the wear scars of DC sputtered Cr–Bcoatings after 250 cycles; a) is a region where the majority of coating hasdelaminated; b) is a region where the majority of coating has survived.

1476 M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

coverage of the coating surface by transferred Al and to debrisgeneration. Afterwards, a steady state is reached due to thecompeting events of Al attachment, detachment, transfer filmformation, debris generation and displacement of material fromthe wear track by the reciprocating sliding ball. At about 2000cycles the coating fails, due to fatigue at the coating–substrateinterface. The COF beyond this point reduces to a value typical ofaluminium sliding against itself.

A number of PMS CrB2 coated samples were tested for wearagainst aluminium alloy. All the samples were investigated by themethods discussed above. SEM in the BEI mode turned out to beparticularly useful as it allowed the fractional area of thewear trackcovered by transferred aluminium to be calculated. It was thuspossible to compare the relative performance of the coatings inthat respect. The results of this comparison are presented in Fig. 7.

Careful analysis of CrB2 coatings by SEM in BEI moderevealed that all CrB2 coatings, despite different depositionparameters, microstructure and mechanical properties performedvery similarly. That is the area covered by aluminium was alwayssmall, in the range of 3.5 to 4.7%. The average value of the areacovered by aluminium for wear tracks of PMS CrB2 coating wascalculated to be 4.1±1.2% (Fig. 7). The individual CrB2 samplesare, therefore, not identified in Fig. 7. Rather, the results for thesamples which exhibited minimum and maximum coverage byaluminium are presented, as well as the average value.

Analysis of TiN and CrN coated, and bare polished, AISI M2tool steel samples was carried out for comparison, revealing thatthe area covered by aluminium in these samples was 5.6, 12.4 and

18.7%, respectively (Fig. 7). The results, therefore, demonstrate asignificantly reduced tendency of PMS CrB2 coatings to pick-upaluminium as compared to CrN and AISI M2 tool steel. Theoverall results, with the exception of TiN, are also in a goodagreement with those reported by Konca et al. [6] who havestudied coatings such as TiB2, TiN and CrN. Our PMS TiNunexpectedly showed performance almost as good as that of PMSCrB2 coatings. The reasons for the low tendency of our TiN topick upAl are unclear and require further investigation. However,results also reveal that the coating surface coverage by Alcorrelates well with the steady state COF exhibited by coatings.COFs of coatings which had a lower tendency to pick upAl (PMSCrB2 and TiN) were higher ~0.55, while the COFs of sampleswith a higher tendency to pick up Al were ~0.4 — the typicalvalue for aluminium sliding against aluminium. From this it canbe concluded that higher COFs found for PMS CrB2 and TiNcoatings are due to prevention of Al adhesion to the coating (andtransfer film formation), resulting from low chemical affinity.

3.3. Reciprocating sliding wear testing against AISI 52100 steel

The friction curves obtained for coatings tested against AISI52100 steel at 5 N normal load, 0.1 ms−1 sliding speed underambient conditions (20–30% humidity; 20 °C) are given in aFig. 8. The performance of PMS CrB2 coatings is, again,compared to that of TiN and CrN coatings. Before failure, atabout 2500 cycles, CrB2 coatings exhibit a stable COF of ~0.6.The relatively early failure of CrB2 coatings is explained by thefact that, in contrast to CrN and TiN coatings, they did not havea metallic adhesion interlayer and are, therefore, moresusceptible to fatigue failure at the coating-substrate interface.The CrN coating exhibited a slightly higher coefficient offriction than CrB2 starting at ~0.6 and then stabilising at ~0.7.The TiN coating, after the initial 400 cycles during which theCOF was 0.5, exhibited quite a low COF of about 0.45 for thefollowing ~3000 cycles. This result is an improvement for PVDTiN, which, under testing conditions similar to the ones used inthis study, usually exhibits COFs of ~0.6 and higher [26–28].

Analysis of wear scars by SEM after an initial 250 cycles wascarried out to study the wear mechanisms of coatings. Fig. 9a–fshows the characteristic features of a CrB2 coating wear scar.Fig. 9b shows regions where the coating has been removed dueto the localised coating adhesion failure and delamination. Theremoved coating is milled by the reciprocating sliding ball andhard CrB2 particles produced in that way contribute to increasedcoating wear. Fig. 9c shows some adhered steel transferred fromball to the coating (composition confirmed by EDX). Steeltransferred from ball to coating undergoes the same processes asdiscussed for Al in Section 3.2 and facilitates the formation of atransfer film. Some scratches induced by abrasive wear andcracks (the consequence of fatigue failure at the coating substrateinterface) are also visible in this figure; they are shown moreclearly at higher magnification in Fig. 9d. Fig. 9a shows theoverall appearance of the wear scar which looks to be relativelyclean, i.e. with not particularly much wear debris generated.Analysis of the debris at higher magnifications revealed thatbesides the brighter (oxide) particles visible in Fig. 9a and e,

Fig. 12. SEM SEI micrographs showing the wear scars of PMS CrN coatingsafter 250 cycles at a) low and b) high magnification.

1477M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

there are less bright, and therefore more difficult to observe, roll-shaped particles (Fig. 9e,f). These particles originate from theoxidised transfer film formed on the surface of the wear scar.

For comparison, wear tracks of DC sputtered (rough andcolumnar) Cr–B coatings after the same initial 250 cycles werealso analysed by SEM (Fig. 10). For these films, wear processessuch as coating removal (due to adhesive failure), debrisgeneration and coating wear rate, were much more pronounced.This is attributed to factors such as the coating roughness,structure and composition and the influence of these on the wearmechanisms. Although a significant part of DC Cr–B coatingwas removed from the wear track (Fig. 10a), there were still afew regions where the coating survived intact (Fig. 10b). Theseregions were used to measure the profiles of the wear tracks.

SEMmicrographs of wear tracks for TiN and CrN coatings areshown in Figs. 11 and 12. It was found that the TiN coating weartrack morphology was significantly different from that of bothCrB2 and CrN coated samples (Figs. 9a and 12). There was muchmore debris generated,which accumulated at the boundaries of thewear scar (Fig. 11a) and the entire surface of the track was coveredby a relatively thick (and therefore easily distinguishable) transferfilm (Fig. 11b). The presence of this thick transfer film and (moreimportantly) its composition and properties, are thought to be thereasons for the low coefficient of friction that TiN exhibited(Fig. 8). Transfer layers are known to have a significant influenceon the tribological behaviour of coatings [29].

Much less debris was generated in the wear track of CrN(Fig. 12a) compared to TiN — but still more than in the case of

Fig. 11. SEM SEI micrographs showing the wear scars of PMS TiN coatingsafter 250 cycles at a) low and b) high magnification.

CrB2. A transfer layer, although clearly much thinner than that ofTiN, could be observed at higher magnifications, and is shown inFig. 12b. From these results, it follows qualitatively that there isa correlation between the properties of the transfer film (such asthickness and composition) and the coefficient of friction that thecoatings exhibited during reciprocating–sliding against AISI52100 steel. That is, the TiN (with its thick transfer film)exhibited a coefficient of friction of 0.42–0.45, while CrB2 andCrN coatings (with their cleaner wear tracks and thinner transferfilms) have exhibited COFs of ~0.6 and 0.6–0.7, respectively.

The wear rate of coatings after 250 cycles was investigated byusing surface profile scans of the wear tracks to calculate theworn area. Pulsed-DC sputtered CrB2 coatings exhibited weartrack cross-section areas from about 10 to 30 mm2. As nocorrelation to the deposition parameters or film properties hasbeen found the samples are not identified. On the other hand, it isimportant to mention here that CrB2 coatings possessing lowerhardness (i.e. ~20 GPa) performed as well as those exhibitinghigh hardness values (~35 GPa), suggesting that high hardnessis not a necessary prerequisite for good anti-wear performance ofthese coatings when sliding against counterface materials suchas AISI 52100 steel. Virtually no wear (i.e. not possible toquantify due on one hand to wear being very low and on the otherhand to relatively more debris generated, as compared to PMSCrB2 coatings) has been found for PMS TiN and CrN coatings.

In Fig. 13, the wear track profiles are compared for the worst(in terms of wear) PMS CrB2 coating (wear track area ~30 mm2)and a DC deposited Cr–B coating (wear track area ~110 mm2). Itfollows from this example that the wear rate of DC deposited Cr–

Fig. 13. Wear scar profiles of (a) PMS CrB2 coating, which exhibited themaximum wear, and (b) DC Cr–B coating.

1478 M. Audronis et al. / Surface & Coatings Technology 202 (2008) 1470–1478

B coatings was some 3.5 times higher than that of even the worstPMS CrB2 coated sample. If the best PMS CrB2 coatings wereconsidered (wear track area ~10mm2) the improvement would betenfold. Incorporation of an adhesion enhancing Cr interlayer intothe CrB2 coated samples would probably reduce significantly thegeneration of hard and abrasive CrB2 containing wear debris andwould therefore lead to further improvement in wear behaviour.

4. Conclusions

Pulsed magnetron sputtered CrB2 and TiN exhibited steadystate COFs ~0.55 and ~0.5 respectively when slid against analuminium alloy ball counterface. These coatings also had thelowest tendency to pick-up Al as compared to CrN and bareAISI M2 tool steel. For CrB2 and TiN coatings respectively,4.1% and 5.6% of the wear track area was covered by Al after250 cycles. PMS CrN, as well as bare AISI M2 tool steel, slidagainst aluminium alloy exhibited steady state COFs ~0.4 — atypical value for Al sliding against Al. CrN and AISI M2 toolsteel also had a higher tendency to pick up Al. For CrN coatedand bare polished AISI M2 steel samples respectively, 12.4%and 18.7% of the wear track area was covered by Al after thesame 250 cycles. PMS CrB2 coatings, therefore, demonstrated asuperior performance with respect to the amount of aluminiumadhered to the wear track. The higher COFs found for PMSCrB2 and TiN coatings were due to prevention of Al adhesion to

the coating and transfer layer formation (due to low chemicalaffinity). PMS CrB2 coatings also exhibited improved frictionbehaviour as compared to the DC sputtered Cr–B films.

When slid against AISI 52100 steel, PMS CrB2, CrN andTiN coatings exhibited coefficients of friction of 0.6, 0.6–0.7and 0.43–0.45, respectively. The tribological behaviour ofcoatings was in this case found to depend strongly on transferfilm formation and its properties. The coating wear rate was upto ten times lower for pulsed magnetron sputtered CrB2 coatingscompared to DC sputtered Cr–B coatings.

Acknowledgments

The assistance of Ms Dawn Bussey in nanoindentationtesting of TiN and CrN films is acknowledged with thanks.

References

[1] M. Berger, S. Hogmark, Wear 252 (2002) 557.[2] M. Berger, S. Hogmark, Surface and Coatings Technology 149 (2002) 14.[3] M. Berger, L. Karlsson, M. Larsson, S. Hogmark, Thin Solid Films 401

(2001) 179.[4] M. Berger, M. Larsson, S. Hogmark, Surface and Coatings Technology

124 (2000) 253.[5] T. Bjork, M. Berger, R. Westergard, S. Hogmark, J. Bergstrom, Surface

and Coatings Technology 146–147 (2001) 33.[6] E. Konca, Y.T. Cheng, A.M. Weiner, J.M. Dasch, A. Erdemir, A.T. Alpas,

Surface and Coatings Technology 200 (2005) 2260.[7] C. Mitterer, Journal of Solid State Chemistry 133 (1997) 279.[8] M. Zhou, M. Nose, Y. Makino, K. Nogi, Thin Solid Films 343–344 (1999)

234.[9] C. Pfohl, A. Bulak, K.T. Rie, Surface and Coatings Technology 131 (2000)

141.[10] E. Kelesoglu, C.Mitterer, Surface and Coatings Technology 98 (1998) 1483.[11] M. Audronis, P.J. Kelly, R.D. Arnell, A. Leyland, A. Matthews, Surface

and Coatings Technology 200 (2005) 1366.[12] M. Audronis, P.J. Kelly, R.D. Arnell, A.V. Valiulis, Surface and Coatings

Technology 200 (2006) 4166.[13] S. Schiller, K. Goedicke, J. Reschke, V. Kirchhoff, S. Schneiderm, F.

Milde, Surface and Coatings Technology 61 (1993) 331.[14] M. Audronis, P.J. Kelly, R.D. Arnell, A. Leyland, A. Matthews, Surface

and Coatings Technology 200 (2005) 1616.[15] M. Audronis, A. Leyland, P.J. Kelly, A. Matthews, Surface and Coatings

Technology 201 (2006) 3970.[16] M. Audronis, P.J. Kelly, A. Leyland, A. Matthews, Thin Solid Films 515

(2006) 1511.[17] C.Rodenburg,W.M.Rainforth, Journal ofApplied Physics 100 (2006) 114902.[18] K.L. Scrivener, Cement and Concrete Composites 26 (2004) 935.[19] R. Yang, N.R. Buenfeld, Cement and Concrete Research 31 (2001) 437.[20] P. Roschger, P. Fratzl, J. Eschberger, K. Klaushofer, Bone 23 (1998) 319.[21] M. Audronis, P.J. Kelly, A. Leyland, A. Matthews, Journal of Physics:

Conference Series 26 (2006) 355.[22] P.B. Barna, M. Adamik, Thin Solid Films 317 (1998) 27.[23] R. Messier, A.P. Giri, R.A. Roy, Journal of Vacuum Science & Technology.

A. Vacuum, Surfaces, and Films 2 (1984) 500.[24] I. Petrov, P.B. Barna, L. Hultman, J.E. Greene, Journal of Vacuum Science

& Technology. A. Vacuum, Surfaces, and Films 21 (2003) S117.[25] J. Tang, L. Feng, J.S. Zabinski, Surface and Coatings Technology 99

(1998) 242.[26] I.L. Singer, S. Fayeulle, P.D. Ehni, Wear 149 (1991) 375.[27] Z.P. Huang, Y. Sun, T. Bell, Wear 173 (1994) 13.[28] J. Takadoum, H. Houmid Bennani, Surface and Coatings Technology 96

(1997) 272.[29] I.L. Singer, S.D. Dvorak, K.J. Wahl, T.W. Scharf, Journal of Vacuum

Science & Technology. A. Vacuum, Surfaces, and Films 21 (2003).

top related